Hunyuan Image 3.0 is Tencent's open-weight text-to-image model — an 80-billion-parameter mixture-of-experts model Tencent open-sourced in September 2025, built for long, detailed prompts and accurate text rendering in both English and Chinese. Uncutly.AI added it to the generate page's model picker, uncensored for verified adult creators 18+.
Generate with Hunyuan ImageHunyuan Image is live today on Uncutly's generate page — open the model picker and select it directly. It doesn't have a curated one-click template yet, so for now it's a build-your-own-prompt model; that catalog is expanding.
Hunyuan Image 3.0 accepts prompts of over 1,000 characters and is tuned to follow multi-requirement instructions and render in-image text accurately — useful when a scene has several specific details that all need to show up.
Hunyuan Image 3.0 costs 359 credits per image on Uncutly, roughly 3x FLUX.1 Kontext Pro (119) and nearly 8x Z-Image (45). That premium tracks its size: at 80B total parameters, it is a materially larger model than anything else in the image lineup, and the credit cost passes that inference cost through directly. Worth reaching for on prompts that actually need its long-context and text-rendering strengths, not as a default pick.
We could not find an independent third-party arena result specific to Hunyuan Image 3.0 differentiated from Tencent's own release claims at the time of writing — most public comparisons we found benchmark GLM-Image or Qwen-Image variants directly against each other rather than against Hunyuan Image 3.0 head-to-head. Treat Tencent's own text-rendering and world-knowledge claims as vendor-reported until independent arena data specifically covering this model is available.
Yes — select it directly from the model picker on the generate page. It does not yet have a curated one-click template.
Yes, for verified adult creators 18+, subject to Uncutly's Content Policy.
80 billion total parameters using a mixture-of-experts architecture, among the largest open-source text-to-image models released to date.
Both handle multi-requirement prompts well; Hunyuan Image additionally supports much longer prompts (1,000+ characters) and stronger bilingual text rendering.
At 359 credits it is the most expensive image model on Uncutly, reflecting its 80-billion-parameter size — roughly 3x FLUX.1 Kontext Pro and nearly 8x Z-Image. Reserve it for prompts that need its long-context or text-rendering strengths.











